
import pandas as pd
import geopandas as gpd
import numpy as np
import os
import glob
import matplotlib.pyplot as plt

import gmaps
import gmaps.datasets

import requests
import re

# jupyter nbconvert --to html --TemplateExporter.exclude_input=True Eesti_haldusjaotus.ipynb --output='EST_regions.html'


# 2018 county recode to NUTS 3
keys = ['Harju maakond',
        'Hiiu maakond','Lääne maakond','Pärnu maakond','Saare maakond',
        'Järva maakond','Lääne-Viru maakond','Rapla maakond',
        'Ida-Viru maakond',
        'Jõgeva maakond','Põlva maakond','Tartu maakond','Valga maakond','Viljandi maakond','Võru maakond']

values = ['North',
         'West','West','West','West',
         'Central','Central','Central',
         'N_East',
         'South','South','South','South','South','South']

nuts_dict = {key: [] for key in keys}
for k, v in zip(keys, values):
    nuts_dict[k].append(v)


# current county map
mknd2021 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/2021/maakond_2021-01-01.shp')
mknd2021['NUTS'] = mknd2021['MNIMI'].replace(nuts_dict)
mknd2021['MNIMI2'] = mknd2021['MNIMI'].str.split().str.get(0)

fig, ax = plt.subplots(figsize=(20,10)) 
mknd2021.plot(column='NUTS', edgecolor='black', ax=ax, cmap='Pastel2')
ax.set_title("2021 county borders (black), colour-filled by NUTS 3 regions")
mknd2021.apply(lambda x: ax.annotate(text=x.MNIMI2, xy=x.geometry.centroid.coords[0], ha='center', size=16), axis=1)
ax.set_axis_off()


mknd1997 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1997/maakond_1997-12-02.shp')
# ov1997 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1997/omavalitsus_1997-12-02.shp')

fig, ax = plt.subplots(figsize=(20, 10))
q = mknd2021.plot(column='NUTS', cmap='Pastel2', ax = ax)
q = mknd1997.plot(color='none', edgecolor='black', linewidths=2, ax=ax)
# q = ov1997.plot(color='none', edgecolor='red', linewidths=0.5, ax=ax)
q.set_title("1997 counties (black), local gov-s (red) over NUTS 3 (colour-fill)")
mknd2021.apply(lambda x: ax.annotate(text=x.MNIMI2, xy=x.geometry.centroid.coords[0], ha='center', size=16), axis=1)
q.set_axis_off()


rjn1989 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1989/Rajoon_1989.shp')
# kn1989 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1989/Kylanoukogu_1989.shp')

fig, ax = plt.subplots(figsize=(20, 10))
p = mknd2021.plot(column='NUTS', cmap='Pastel2', ax = ax)
p = rjn1989.plot(color='none', edgecolor='black', linewidths=2, ax=ax)
#p = kn1989.plot(color='none', edgecolor='red', linewidths=0.5, ax=ax)
p.set_title("1989 rayons (black), selsoviets (red) over NUTS 3 (colour-fill)")
rjn1989.apply(lambda x: ax.annotate(text=x.NIMI, xy=x.geometry.centroid.coords[0], ha='center', size=16), axis=1)
p.set_axis_off()


rjn1979 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1979/Rajoon_1979.shp')
# kn1979 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1979/Kylanoukogu_1979.shp')

fig, ax = plt.subplots(figsize=(20, 10))
q = mknd2021.plot(column='NUTS', cmap='Pastel2', ax = ax)
q = rjn1979.plot(color='none', edgecolor='black', linewidths=2, ax=ax)
# q = kn1979.plot(color='none', edgecolor='red', linewidths=0.5, ax=ax)
q.set_title("1979 rayons (black), selsoviets (red) over NUTS 3 (colour-fill)")
rjn1979.apply(lambda x: ax.annotate(text=x.NIMI, xy=x.geometry.centroid.coords[0], ha='center', size=16), axis=1)
q.set_axis_off()


rjn1959 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1959/Rajoon_1959.shp')
# kn1959 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1959/Kylanoukogu_1959.shp')

fig, ax = plt.subplots(figsize=(20, 10))
p = mknd2021.plot(column='NUTS', cmap='Pastel2', ax = ax)
p = rjn1959.plot(color='none', edgecolor='black', linewidths=2, ax=ax)
# p = kn1959.plot(color='none', edgecolor='red', linewidths=0.5, ax=ax)
p.set_title("1959 rayons (black), selsoviets (red) over NUTS 3 (colour-fill)")
rjn1959.apply(lambda x: ax.annotate(text=x.NIMI, xy=x.geometry.centroid.coords[0], ha='center', size=12), axis=1)
p.set_axis_off()


# 1959 rayon recode to NUTS3
keys3 = ['Harju','Keila','Tallinn',
        'Hiiumaa','Haapsalu','Lihula','Pärnu','Kingissepa','Vändra',
        'Märjamaa','Rapla','Paide','Tapa','Väike-Maarja','Rakvere',
        'Narva','Kohtla-Järve','Jõhvi','Sillamäe',
        'Jõgeva','Põltsamaa','Põlva','Tartu','Elva','Räpina','Valga','Viljandi','Võru','Abja']

values3 = ['North','North','North',
         'West','West','West','West','West','West',
         'Central','Central','Central','Central','Central','Central',
         'N_East','N_East','N_East','N_East',
         'South','South','South','South','South','South','South','South','South','South']

rayon_dict = {key: [] for key in keys3}
for k, v in zip(keys3, values3):
    rayon_dict[k].append(v)


mknd1938 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1938/Maakond_1938.shp')
# vald1938 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1938/Vald_1938.shp')

fig, ax = plt.subplots(figsize=(20, 10))
q = mknd2021.plot(column='NUTS', cmap='Pastel2', ax = ax)
# q = vald1938.plot(color='none', edgecolor='red', linewidths=0.5, ax=ax)
q = mknd1938.plot(color='none', edgecolor='black', linewidths=2, ax=ax)
q.set_title("1938 counties (black), parishes (red) over NUTS 3 (colour-fill)")
mknd1938.apply(lambda x: ax.annotate(text=x.NIMI, xy=x.geometry.centroid.coords[0], ha='center', size=16), axis=1)
q.set_axis_off()


mknd1922 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1922/Maakond_1922.shp')
# vald1922 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1922/Vald_1922.shp')

fig, ax = plt.subplots(figsize=(20, 10))
p = mknd2021.plot(column='NUTS', cmap='Pastel2', ax = ax)
# p = vald1922.plot(color='none', edgecolor='red', linewidths=0.5, ax=ax)
p = mknd1922.plot(color='none', edgecolor='black', linewidths=2, ax=ax)
p.set_title("1922 counties (black), parishes (red) over NUTS 3 (colour-fill)")
mknd1922.apply(lambda x: ax.annotate(text=x.NIMI, xy=x.geometry.centroid.coords[0], ha='center', size=16), axis=1)
p.set_axis_off()


mknd1917 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1917/Maakond_1917.shp')
# khlk1917 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1917/Kihelkond_kuni_1917.shp')
# mois1917 = gpd.read_file('/Users/martin/Documents/Loendused/Kaardid/1917/moisamaad_1917.shp')

fig, ax = plt.subplots(figsize=(26, 12))
p = mknd2021.plot(column='NUTS', cmap='Pastel2', ax = ax)
# p = khlk1917.plot(color='none', edgecolor='red', linewidths=0.5, ax=ax)
p = mknd1917.plot(color='none', edgecolor='black', linewidths=2, ax=ax)
# p = mois1917.plot(color='none', edgecolor='blue', linewidths=0.5, ax=ax)
p.set_title("1917 counties (black), kihelkond (red) over NUTS 3 (colour-fill)")
mknd1917.apply(lambda x: ax.annotate(text=x.NIMI, xy=x.geometry.centroid.coords[0], ha='center', size=16), axis=1)
p.set_axis_off()


from plotly.offline import download_plotlyjs, init_notebook_mode, plot, iplot
from plotly.graph_objs import Scatter, Figure, Layout
import plotly
import plotly.graph_objs as go
import plotly.express as px
from IPython.display import Markdown as md
init_notebook_mode(connected=False)


cens_files = glob.glob('/Users/martin/Documents/Loendused/Stat_ee/*.csv')
cens_files

['/Users/martin/Documents/Loendused/Stat_ee/loendus_1934.csv',
 '/Users/martin/Documents/Loendused/Stat_ee/loendus_1897.csv',
 '/Users/martin/Documents/Loendused/Stat_ee/loendus_1881.csv',
 '/Users/martin/Documents/Loendused/Stat_ee/loendus_1922.csv',
 '/Users/martin/Documents/Loendused/Stat_ee/loendus_1979.csv',
 '/Users/martin/Documents/Loendused/Stat_ee/loendus_1989.csv',
 '/Users/martin/Documents/Loendused/Stat_ee/loendus_1959.csv',
 '/Users/martin/Documents/Loendused/Stat_ee/loendus_1970.csv']


# 1959 census
df_1959 = pd.concat([pd.read_csv(f) for f in cens_files[6:7]], 
               ignore_index=True)

df_1959 = df_1959[pd.to_numeric(df_1959['Unnamed: 0'], errors='coerce').notnull()] # ainult nummerdatud read
df_1959 = df_1959[~df_1959['Rajoon, linn'].isin(['Kogurahvastik','Linnarahvastik','Maarahvastik'])] # rajoonid + linnad
print(df_1959['Kokku'].sum())

1196791.0


df_1959['Rajoon, linn'] = df_1959['Rajoon, linn'].str.title()
df_1959['NUTS3'] = df_1959['Rajoon, linn'].replace(rayon_dict)

df_1959 = df_1959.groupby(['Aasta', 'NUTS3'])['Kokku'].agg('sum').reset_index()


df_1959


import json
import requests
import urllib.request
import pandas as pd
import pyjstat

from pyjstat import pyjstat
from collections import OrderedDict


import pandas as pd
import matplotlib.pyplot as plt
from plotly.offline import download_plotlyjs, init_notebook_mode, plot, iplot
from plotly.graph_objs import Scatter, Figure, Layout
import plotly
import plotly.graph_objs as go
from plotly.subplots import make_subplots
import plotly.express as px
init_notebook_mode(connected=False)


# Population by county 1990-2017

url = "https://andmed.stat.ee/api/v1/en/stat/RV022"

payload = {
  "query": [
    {
      "code": "Maakond",
      "selection": {
        "filter": "item",
        "values": [
          "2", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", "16", "17", "18", "19"
        ]
      }
    },
    {
      "code": "Sugu",
      "selection": {
        "filter": "item",
        "values": [
          "K"
        ]
      }
    },
    {
      "code": "Vanuserühm",
      "selection": {
        "filter": "item",
        "values": [
          "999"
        ]
      }
    }
  ],
  "response": {
    "format": "json-stat2"
  }
}


data = requests.post(url, data=json.dumps(payload))
r = pyjstat.from_json_stat(data.json(object_pairs_hook=OrderedDict))
# print(r)


df = pd.DataFrame(r[0], columns=['Year','County','Sex','Age group','value'])
df['County'] = df.County.str.split(" ").str[0]
df.loc[df.County == 'County', 'County'] = 'Unknown'


dat_edk = pd.read_csv("/Users/martin/Documents/Loendused/EKDK/maakonnad_1965-1990.csv")
dat_edk = dat_edk[:-1]

df2 = pd.concat([dat_edk.melt(id_vars='Year', var_name="County"),df])


# County name recode to NUTS3
keys2 = ['Harju',
         'Hiiu','Lääne','Pärnu','Saare',
         'Järva','Lääne-Viru','Rapla',
         'Ida-Viru',
         'Jõgeva','Põlva','Tartu','Valga','Viljandi','Võru']

values2 = ['North',
           'West','West','West','West',
           'Central','Central','Central',
           'N_East',
           'South','South','South','South','South','South']

nuts_dict2 = {key: [] for key in keys2}
for k, v in zip(keys2, values2):
    nuts_dict2[k].append(v)


df3 = df2[['Year','County','value']].copy()
df3['NUTS3'] = df3['County'].replace(nuts_dict2)

df3 = df3[['Year','value','NUTS3']]

df3 = df3.groupby(['Year', 'NUTS3'])['value'].agg('sum').reset_index()


df_1959 = df_1959.rename({'Aasta': 'Year', 'Kokku': 'value'}, axis=1)

df4 = pd.concat([df3, df_1959])
df4 = df4.rename({'value':'Population'}, axis=1)


px.defaults.width = 1000
px.defaults.height = 500


fig = px.bar(df4, x="Year", y="Population", 
             color="NUTS3", hover_name="NUTS3", log_y=False)
fig.update_layout(title="Population by NUTS3 units 1959-2017")
fig.show()


url = "https://andmed.stat.ee/api/v1/en/stat/RV11"

payload = {
  "query": [
    {
      "code": "Maakond",
      "selection": {
        "filter": "item",
        "values": ["37","39","44","49","51","57","59","65","67","70","74","78","82","84","86","99"]
      }
    }
  ],
  "response": {
    "format": "json-stat2"
  }
}


data = requests.post(url, data=json.dumps(payload))
r = pyjstat.from_json_stat(data.json(object_pairs_hook=OrderedDict))

births = pd.DataFrame(r[0], columns=['Year','County','value'])
births['County'] = births.County.str.split(" ").str[0]
births.loc[births.County == 'County', 'County'] = 'Unknown'
births['Year'] = pd.to_numeric(births['Year'])


births['NUTS3'] = births['County'].replace(nuts_dict2)

births = births.groupby(['Year', 'NUTS3'])['value'].agg('sum').reset_index()
births = births.rename({'value':'Births'}, axis=1)

fig = px.bar(births, x="Year", y="Births", 
             color="NUTS3", hover_name="NUTS3", log_y=False)
fig.update_layout(title="Number of births")
fig.show()


url = "https://andmed.stat.ee/api/v1/en/stat/RV52"
payload = {
  "query": [
    {
      "code": "Sugu",
      "selection": {
        "filter": "item",
        "values": ["1"]
      }
    },
    {
      "code": "Asustusüksuse liik",
      "selection": {
        "filter": "item",
        "values": ["1"]
      }
    },
    {
      "code": "Elukoht",
      "selection": {
        "filter": "item",
        "values": ["H2","37","39","44","49","51","57","59","65","67","70","74","78","82","84","86"]
      }
    },
    {
      "code": "Vanuserühm",
      "selection": {
        "filter": "item",
        "values": ["2"]
      }
    }
  ],
  "response": {"format": "json-stat2"}
}


data = requests.post(url, data=json.dumps(payload))
r = pyjstat.from_json_stat(data.json(object_pairs_hook=OrderedDict))

imr = pd.DataFrame(r[0], columns=['Year','Place of residence','value'])
imr = imr.rename({'Place of residence': 'County'}, axis=1)
imr['County'] = imr.County.str.split(" ").str[0]
imr.loc[imr.County == 'Place', 'County'] = 'Unknown'


dat_edk_imr = pd.read_csv("/Users/martin/Documents/Loendused/EKDK/infant_deaths_1965-1990.csv")
dat_edk_imr = dat_edk_imr[:-2]


imr3 = pd.concat([dat_edk_imr.drop('Total',axis=1).melt(id_vars='Year', var_name="County"), imr])

# imr3 = imr2[['Year','County','value']].copy()
imr3['NUTS3'] = imr3['County'].replace(nuts_dict2)

#imr3 = imr3[['Year','value','NUTS3']]

imr3 = imr3.groupby(['Year', 'NUTS3'])['value'].agg('sum').reset_index()
imr3 = imr3.rename({'value':'Inf.Deaths'}, axis=1)

imr3['Year'] = pd.to_numeric(imr3['Year'])

fig = px.bar(imr3, x="Year", y="Inf.Deaths", 
             color="NUTS3", hover_name="NUTS3", log_y=False)
fig.update_layout(title="Number of infant deaths")
fig.show()


# Infant mortality rate
imr4 = pd.merge(imr3, births, on=['NUTS3','Year'])

imr4['IMR'] = imr4['Inf.Deaths']/imr4['Births']*1000

fig = px.line(imr4, x="Year", y="IMR", color="NUTS3", hover_name="NUTS3", symbol="NUTS3")
fig.update_layout(title="Infant mortality rate")
fig.show()


# Regional GDP 1995-2020

#Post Request
url = "https://andmed.stat.ee/api/v1/en/stat/RAA0051"

payload = {
  "query": [
    {
      "code": "Maakond",
      "selection": {
        "filter": "item",
        "values": ["37","39","44","49","51","57","59","65","67","70","74","78","82","84","86"]
      }
    },
     {
      "code": "Komponent/Majandussektor",
      "selection": {
        "filter": "item",
        "values": [
          "1",
          "2",
          "3"
        ]
      }
    },
    {
      "code": "Näitaja",
      "selection": {
        "filter": "item",
        "values": [
          "1"
        ]
      }
    }
  ],
  "response": {
    "format": "json-stat2"
  }
}


data = requests.post(url, data=json.dumps(payload))
results = pyjstat.from_json_stat(data.json(object_pairs_hook=OrderedDict))
# print(results)


df = pd.DataFrame(results[0], columns=['Year','County','Component/ Economic sector','value'])
df = df.rename({'Component/ Economic sector':'Sector'}, axis=1)
df.tail()


# px.defaults.template = "ggplot2"
# px.defaults.color_continuous_scale = px.colors.sequential.Blackbody
px.defaults.width = 1000
px.defaults.height = 700


df.Sector.value_counts()


df['NUTS3'] = df['County'].str.split().str.get(0)
df['NUTS3'] = df['NUTS3'].replace(nuts_dict2)
df1 = df.groupby(['Year', 'NUTS3','Sector'])['value'].agg('sum').reset_index()
fig = px.line(df1, x="Year", y="value", color="NUTS3", hover_name="NUTS3", facet_col="Sector", log_y=True)
fig.show()


# Regional GDP 1995-2020

#Post Request
url = "https://andmed.stat.ee/api/v1/en/stat/RAA0053"

payload = {
  "query": [
    {
      "code": "Piirkond",
      "selection": {
        "filter": "item",
        "values": [
          "2",
          "3",
          "4",
          "5",
          "6"
        ]
      }
    },
    {
      "code": "Näitaja",
      "selection": {
        "filter": "item",
        "values": [
          "3"
        ]
      }
    }
  ],
  "response": {
    "format": "json-stat2"
  }
}


data = requests.post(url, data=json.dumps(payload))
results = pyjstat.from_json_stat(data.json(object_pairs_hook=OrderedDict))
# print(results)

df = pd.DataFrame(results[0], columns=['Year','Region','Indicator','value'])
df.tail()


fig = px.bar(df, x="Year", y="value", color="Region")
fig.show()


emp79 = pd.read_csv("/Users/martin/Documents/Loendused/EKDK/tooala_79.csv")
emp89 = pd.read_csv("/Users/martin/Documents/Loendused/EKDK/tooala_89.csv")

emp = pd.concat([emp79, emp89])
emp = emp[emp.NRTAK!='Nonactive']

print(emp['NRTAK'].unique())

['Industry' 'Agriculture' 'Transport' 'Communication' 'Construction'
 'Trade' 'OMP' 'Services' 'Healthcare' 'Education' 'PubAdmin' 'Unknown']


emp2 = emp.melt(id_vars=['NRTAK','Year'], var_name="County")
emp2 = emp2[emp2['County']!='Total']
emp2['NUTS3'] = emp2['County'].replace(nuts_dict2)
emp2 = emp2.groupby(['NUTS3','NRTAK','Year'])['value'].agg('sum').reset_index()

fig = px.bar(emp2.sort_values('value'), x="Year", y="value", color="NRTAK", facet_col='NUTS3',
            color_discrete_sequence=px.colors.qualitative.Alphabet)
fig.update_layout(
    title = 'Number of employees by NRTAK & NUTS3', yaxis_title="Count")
fig.update_xaxes(type='category')
fig.show()


ISIC_to_NRTAK_dict = dict({'Põllumajandus - Agriculture':   'Agriculture', 
                      'Tööstus - Industry':                 'Industry',
                      'Ehitus - Construction':              'Construction',
                      'Transport - Transportation':         'Transport',
                      'Haridus - Education':                'Education',
                      'Kaubandus - Trade':                  'Trade',
                      'Tervishoid - Health care':           'Healthcare',
                      'Muud teenused - Other services':     'Services',
                      'Avalik teenistus - Public service':  'PubAdmin',
                      'Kinnisvara ja äri - Real estate and business': 'Services',
                      'Kaevandus - Mining':                 'Industry',
                      'Eramajapidamised - Private household': 'OMP',
                      'Hotellid-restoranid - Hotels-restaurants': 'Services',
                      'Elektri-veevarustus - Electricity and water supply': 'Services',
                      'Kalandus - Fishing':                 'Agriculture',
                      'Rahandus - Finance':                 'Services'})


pd.set_option("display.max_rows", None)


dat = pd.read_stata('/Users/martin/Documents/Loendused/Tulu-uuringud/tulu.dta')
dat = dat.loc[dat.aasta > 1958]

#dict_raj = dict({1:'Harju', 2:'Hiiu', 3: 'Ida-Viru', 4:'Jõgeva', 5:'Järva', 6:'Lääne', 7:'Lääne-Viru', 8:'Põlva', 
#                 9:'Pärnu', 10:'Rapla', 11:'Saare', 12:'Tartu', 13:'Valga', 14:'Viljandi', 15:'Võru'})


dat['T89'] = dat['a0ter89']
dat['T89'] = dat['T89'].astype('Int64').apply('{:0>4}'.format)
dat['T89'] = pd.to_numeric(dat['T89'].astype(str).str[:2]).astype('Int64').replace(dict_raj)


dat.loc[dat['T89']==0, 'T89'] = dat['a0ter59'].str.split().str.get(0)

dat['NUTS3'] = dat['T89'].replace(nuts_dict2)
dat['NUTS3'] = dat['NUTS3'].replace(rayon_dict)
dat = dat.drop(dat[(dat['NUTS3'].isin([90,99,'Nuia']))].index)
dat.aasta.value_counts()

1975    6562
1981    5672
Name: aasta, dtype: int64


dat = dat.drop(dat[(dat['sektor'].isin(['NA']))].index)
# pd.crosstab([dat.aasta,dat.NUTS3], dat.sektor)
dat.sektor.value_counts()

Tertsiaar - Tertiary     4566
Sekundaar - Secondary    4179
Primaar - Primary        3484
NA                          0
Name: sektor, dtype: int64


dat['NRTAK'] = dat['ISIC'].replace(ISIC_to_NRTAK_dict)
# pd.crosstab(dat['NRTAK'],dat['sektor'])


dat0 = dat.groupby(['NRTAK','aasta','NUTS3'])['ttasu'].agg(['mean']).reset_index().rename({'aasta':'Year'}, axis=1)
# print(dat0.pivot_table(index=['NRTAK'], columns=['Year','NUTS3'],values='mean').round(0))

dat0 = dat[dat.aasta==1981].groupby(['NRTAK','NUTS3'])['ttasu'].agg(['mean']).reset_index()
dat0.pivot_table(index=['NRTAK'], columns=['NUTS3'],values='mean').round(0)


emp2['Year'] = emp2['Year'].astype('category')
dt['total'] = dt['value']*dt['mean']*12/1000_000
dt.head()

dt0 = dt.groupby(['Year','NUTS3'])['total'].agg(['sum']).reset_index()
dt0.pivot_table(index=['NUTS3'], columns=['Year'],values='sum',aggfunc= 'sum',margins=True).round(0).iloc[:,:-1]


dat1 = dat.groupby(['NRTAK','NUTS3','aasta'])['ttasu'].agg(['mean']).reset_index()
dat1 = dat1.dropna()
dat1

fig = px.bar(dat1.sort_values('mean'), x="NUTS3", y="mean", color="NRTAK", facet_col='aasta', barmode='group')
# ,color_discrete_sequence=px.colors.qualitative.Rainbow)
fig.show()


import statsmodels.formula.api as sm


# dat['logttasu'] = np.log(dat.ttasu)
model = sm.ols('ttasu ~ NRTAK + NUTS3 + C(aasta)', data=dat) #  + sugu + haridus
res = model.fit()
print(res.summary())

                            OLS Regression Results                            
==============================================================================
Dep. Variable:                  ttasu   R-squared:                       0.137
Model:                            OLS   Adj. R-squared:                  0.136
Method:                 Least Squares   F-statistic:                     139.0
Date:                Tue, 14 Jun 2022   Prob (F-statistic):               0.00
Time:                        08:39:06   Log-Likelihood:                -71219.
No. Observations:               12229   AIC:                         1.425e+05
Df Residuals:                   12214   BIC:                         1.426e+05
Df Model:                          14                                         
Covariance Type:            nonrobust                                         
=========================================================================================
                            coef    std err          t      P>|t|      [0.025      0.975]
-----------------------------------------------------------------------------------------
Intercept               191.1168      2.391     79.917      0.000     186.429     195.804
NRTAK[T.Construction]    15.6540      2.929      5.345      0.000       9.913      21.395
NRTAK[T.Education]      -52.4408      3.274    -16.019      0.000     -58.858     -46.024
NRTAK[T.Healthcare]     -72.9940      3.858    -18.920      0.000     -80.556     -65.432
NRTAK[T.Industry]       -20.1491      2.174     -9.267      0.000     -24.411     -15.887
NRTAK[T.OMP]            -64.9256      6.798     -9.551      0.000     -78.250     -51.601
NRTAK[T.PubAdmin]       -41.4516      4.856     -8.536      0.000     -50.970     -31.933
NRTAK[T.Services]       -60.5556      2.909    -20.820      0.000     -66.257     -54.854
NRTAK[T.Trade]          -61.0049      3.423    -17.824      0.000     -67.714     -54.296
NRTAK[T.Transport]      -14.4734      2.977     -4.862      0.000     -20.309      -8.638
NUTS3[T.N_East]         -16.7912      2.918     -5.754      0.000     -22.511     -11.071
NUTS3[T.North]           -8.9463      2.535     -3.530      0.000     -13.915      -3.978
NUTS3[T.South]          -11.0883      2.466     -4.496      0.000     -15.923      -6.254
NUTS3[T.West]            -9.7210      2.715     -3.580      0.000     -15.044      -4.398
C(aasta)[T.1981]         36.4765      1.502     24.278      0.000      33.531      39.422
==============================================================================
Omnibus:                     3768.709   Durbin-Watson:                   1.958
Prob(Omnibus):                  0.000   Jarque-Bera (JB):            15196.571
Skew:                           1.483   Prob(JB):                         0.00
Kurtosis:                       7.586   Cond. No.                         11.9
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.

NUTS3	Central	N_East	North	South	West
NRTAK
Agriculture	245.0	209.0	215.0	218.0	229.0
Construction	240.0	214.0	186.0	224.0	226.0
Education	155.0	159.0	170.0	152.0	143.0
Healthcare	122.0	128.0	160.0	135.0	128.0
Industry	190.0	200.0	203.0	199.0	193.0
PubAdmin	123.0	218.0	184.0	138.0	160.0
Services	146.0	133.0	169.0	149.0	148.0
Trade	179.0	127.0	144.0	163.0	152.0
Transport	225.0	204.0	178.0	224.0	196.0

Year	1979	1989
NUTS3
Central	198.0	213.0
N_East	270.0	278.0
North	666.0	744.0
South	437.0	480.0
West	211.0	231.0
All	1782.0	1947.0

Historical administrative units and NUTS 3 in Estonia¶

2021 county borders and NUTS 3 regions¶

Historical borders¶

1997 counties and local governmental units¶

1989 rayons and selsoviets¶

1979 rayons and selsoviets¶

1959 rayons and selsoviets¶

1938 counties and parishes¶

1922 counties and parishes¶

1917 counties and Kirchspiel¶

Population counts¶

1959 census data¶

Stat.Office data 1990-2017¶

Reconstructed 1965-1990 population counts and events¶

Births¶

Infant deaths¶

GDP post-1990¶

Regional employment by industry 1979 & 1989¶

NRTAK & ISIC¶

Net wage income by NRTAK and NUTS3¶

	Aasta	NUTS3	Kokku
0	1959	Central	154360.0
1	1959	N_East	133093.0
2	1959	North	353919.0
3	1959	South	392337.0
4	1959	West	163082.0